Machine learning preflight printing system and methods

Abstract

A printing system includes a printing device. The printing system also includes a preflight system that checks incoming print jobs for possible errors or issues before commencing printing operations. The preflight system implements a generative adversarial network to facilitate the identification of possible problems with printing. The generative adversarial network includes a generative neural network and a discriminatory neural network. The generative neural network introduces errors into input data to train the discriminatory neural network in identifying problems with print jobs. The discriminatory neural network backpropagates its output to train the generative neural network.

1 . A method for performing a check on a print job, the method comprising:
 normalizing data within the print job and metadata associated with the print job;   inputting the normalized data into a neural network, wherein the neural network includes an input layer having input nodes to receive inputs of the normalized data, at least one hidden layer having neural nodes, and an output layer having output nodes;   processing the normalized data within the at least one hidden layer using the neural nodes, wherein each neural node multiplies inputs from a previous node by a weight, sums the multiplied inputs, and feeds the summed inputs to an activation function;   determining outputs for the output nodes from the neural nodes of the at least one hidden layer; and   determining whether the print job includes an issue based on the output of the output nodes.   
     
     
         2 . The method of  claim 1 , wherein normalizing data includes converting the print job into pre-set dictionaries. 
     
     
         3 . The method of  claim 2 , wherein the pre-set dictionaries are fixed sized. 
     
     
         4 . The method of  claim 1 , wherein normalizing data includes converting text within the print job into normalized text. 
     
     
         5 . The method of  claim 1 , further randomizing the inputs of the normalized data. 
     
     
         6 . The method of  claim 1 , further comprising feeding raster image processing performance data as inputs to the input layer. 
     
     
         7 . The method of  claim 1 , further comprising feeding fidelity data as inputs to the input layer. 
     
     
         8 . The method of  claim 1 , wherein the neural network is a discriminatory network. 
     
     
         9 . The method of  claim 8 , further comprising using a generative network to train the discriminatory network. 
     
     
         10 . A computer system for checking a print job prior to printing operations, the computer system comprising:
 a computer processor;   a computer readable storage device, wherein stored program instructions on the computer readable storage device for execution by the computer processor, the stored program instructions configuring the processor to   normalize data within the print job and metadata associated with the print job;   input the normalized data into a neural network, wherein the neural network includes an input layer having input nodes to receive inputs of the normalized data, at least one hidden layer having neural nodes, and an output layer having output nodes;   process the normalized data within the at least one hidden layer using the neural nodes, wherein each neural node multiplies inputs from a previous node by a weight, sums the multiplied inputs, and feeds the summed inputs to an activation function;   determine outputs for the output nodes from the neural nodes of the at least one hidden layer; and   determine whether the print job includes an issue based on the output of the output nodes.   
     
     
         11 . The computer system of  claim 10 , wherein the stored program instructions further configure the processor to convert the print job into pre-set dictionaries. 
     
     
         12 . The computer system of  claim 10 , wherein the pre-set dictionaries are fixed sized. 
     
     
         13 . The computer system of  claim 10 , wherein the stored program instructions further configure the processor to convert text within the print job into normalized text. 
     
     
         14 . The computer system of  claim 10 , wherein the neural network is a discriminatory network. 
     
     
         15 . The computer system of  claim 14 , wherein the stored program instructions further configure the processor to use a generative network to train the discriminatory network. 
     
     
         16 . A method for using a neural network for use in checking printing operations at a printing device, the method comprising:
 normalizing data within a print job and metadata associated with the print job;   capturing raster image processing (RIP) performance data generated by processing the print job;   capturing fidelity data generated while printing the print job within the printing device;   inputting the normalized data, the RIP performance data, and the fidelity data to the neural network; and   determining a file modification to the print job based on an output of the neural network.   
     
     
         17 . The method of  claim 16 , further comprising determining whether the file modification resolved the error. 
     
     
         18 . The method of  claim 16 , further comprising training the neural network using the normalized data. 
     
     
         19 . The method of  claim 16 , wherein the neural network is a discriminatory network. 
     
     
         20 . The method of  claim 19 , further comprising using a generative network to train the discriminatory network.